In the field of digital communication, it is desired to have wide dynamic range and precision for digital numeric representation. However, quantization and saturation effects, typically due to the finite precision resulting from the number of bits chosen to represent a number, may limit dynamic range and accuracy. Increasing the number of bits may improve quantization and saturation effects, but such an increase typically increases the memory requirement and complexity of a system. Accordingly, the processing of communication signals with reasonable quantization and saturation performance, together with a reduced number of bits, is desired.